Hi @MaximMoinat ,
I also drafted a couple of examples, look here.
The structure is the same and I have a couple of suggestions regarding the mapping:
-
I added the mappings to Standard Question-Anwer combinations, where some of the targets are UKB concepts by themselves (no Standard equivalent exist).
-
Qualifier Value concepts (No evidence of, History of) are not intended to be used for event coding - they’re just attributes. How about these: Clinical finding absent, History of clinical finding in subject, Past history of procedure and their new descendants with time context.
-
I would not map to the event such things as Urea correction reason. Or at least don’t specify the reason in the operator concept_id field - it’s off-label usage
-
For Alcohol intake frequency, some may be mapped to the pre-coordinated events: Alcohol consumption unknown, Occasional drinker, Current drinker, etc.
-
We need to be very specific on the presence of the historical context in the question - we review both name and note fields provided by the source.
It’s still not clear how to handle this numeric stuff (\d+.?\d*).
In general, it looks very good, but we will stick with concatenated question-answer pairs for some time.